“Novel graphs for rainfall shortages”

I display my drought poster

Garry’s drought poster

Drought poster for conference

Grasslands Conference Poster Display

A conference paper about drought graphs

This blog “Climate by Surly” has analyses of the current drought at Manilla, mainly in the form of graphs. I was invited to present these findings as a “Poster Paper” at a conference of the Grassland Society of NSW at Gunnedah in July 2019. The poster itself is shown in the photo and the reduced image above.

The paper

The paper is published in two parts: the colour poster shown above, and four pages of text and figures. Both have the title “Novel graphs show extreme rainfall shortages”.
The point of my paper was that these novel graphs bring out features of the current drought at Manilla in a way that could be applied at other times and places.
Because the paper is brief, I have referred the reader to further information in posts in this blog. The construction of each of the three kinds of graph is fully explained in the posts that I linked:
/2019/03/31/march-rain-leaves-drought-extreme/
/2019/04/09/rainfall-shortage-sequence-03-2019/
/2019/04/15/rainfall-shortage-history-manilla/
/2019/04/16/rainfall-shortage-jan-2000-mar-2019/

The complete reference to the paper is:
Speight JG (2019) Novel graphs show extreme rainfall shortages. In
‘Proceedings of the 31st Conference of the Grassland Society of NSW Inc.’ (Eds SR Murphy, SP Boschma, and M Simpson). pp. 36-39. (Grassland Society of NSW Inc., Orange).

The text can be accessed here as a pdf:

Click to access 6-2019-grassland-conf-proc-speight-pp35-39-web-ver.pdf

The paper will become accessible on the website of the Grassland Society of NSW.

The Poster on display

Thanks to the generosity and interest of the proprieters, the original of the poster can be seen at Molly May’s Coffee Shop in the main street of Manilla.

Rainfalls: 8 in the 0.1th percentile

Rainfall status October November 2019

Normal rainfall in November

Rainfall in November 2019 (40.2 mm) was near normal. That reduced shortages at durations from 2-months to 5-months. Otherwise it had little effect: fifteen of twenty longer durations remained extreme shortages.

Values plotted in the 0.1th percentile

For simplicity, the bottom line is labelled with the 0.1th percentile value, and percentile values below 0.1% are plotted on the line. As there are 1600 months of record, both the 2nd-driest month (percentile value 0.063%) and the driest month, (percentile value 0.000%, by convention), which would plot below the line, are plotted on it.
Driest records have again been broken at durations of 12-months (270 mm), 24-months (611 mm), 30-months (834 mm), 72-months (2927 mm), and 96-months (4205 mm). The 96-month record had stood at 4405 mm since November 1919.
A value that equals an earlier record occurs at 84-months (3555 mm).
Values that are 2nd-driest occur at 36-months (1129 mm) and at 240-months (11816 mm). The 240-month (20-year) total is now only 50 mm more than the lowest-ever value of 11766 mm set in February 1931.

How to read the graph

This graph shows all the present rainfall shortages at Manilla, short term and long term, as percentile values. The latest values, as at the end of October 2019, are shown by a thick black line with large circles. Those from one month earlier are shown by a thinner line with small diamonds. [The method is described in “Further Explanation” below.]


Further Explanation

The following notes explain aspects of this work under these listed headings:

Data analysis

Cumulative rainfall totals
Percentile values
Severity of rainfall shortages

Limitations of this analysis

Monthly rainfalls form a single population
Observations are not retrospective
The rain gauge failed

Data analysis

Continue reading

How much rainfall is now lost?

Rainfall deficits in mm at Manilla NSW, Oct 2019

This graph shows the total rainfall shortage in this drought at Manilla NSW. It uses the same rainfall figures as in the October Rainfall Status graph, that is, the rainfall totals for up to 360 months. In this case, I have subtracted from them the normal rainfall totals. That shows the rainfall deficit: how much less rainfall we have had than we normally would.

Generally a 1-year deficit

In many cases, the shortage now is about 650 mm, or one year of normal rainfall. This is true when calculated for 24-months and for several other durations including 360-months. In broad terms, this drought has left Manilla about one year short of rain.

Deficits are smaller at short durations.

A one-year deficit could occur at a duration of 12-months only if there were no rain. A curve on the graph shows “No rain at all”. A second curve is labelled “50% of normal Rain”. These October data show that the rainfall totals for durations below 24-months are around 30% of normal.

Deficits larger than 650 mm occur at some long durations

There are three dips in the curve where deficits are greater than one year of rainfall.

  1. The 3-year deficit is now 820 mm, due to very low rainfall in the last six months. Such values may yet appear at other durations.
  2. The 7-year deficit is now 920 mm. This deficit is due to the inclusion of severe rainfall shortages of 12-month to 30-month duration in 2013 and 2014. These appear on this drought duration graph, but were scarcely noticed at the time.
  3. The 20-year deficit is now 1100 mm. The 650 mm deficit due to the current drought is supplemented by deficits of some 200 mm carried forward from both the 2013-14 shortages (above) and the extreme 12-month shortage of 2002.

What the deficits mean

In times of normal or excess rainfall, the whole terrain is kept supplied with water. It goes to aquifers deep underground, to shallow aquifers near streams, to stream-flow, and to the soil and plants. In general, the longer the duration of normal or excess rainfall, the larger the reservoirs, both underground and on the surface, that can be filled.
The present deficit of one year of rainfall must have drawn down all the reserves of water in the terrain. The empty state of the Keepit Dam (0.8%) is an obvious result, but all other reserves must have fallen.

21stC Rain Shortage Oct ’19

Record of rainfall shortages Jan 2000 Oct 2019

Since the graph for May, the drought at Manilla has become more severe at all durations, from two months up to 240 months.

[For explanation of this graph, see below: “About drought duration graphs”.]

Shortages Jan 2000 to Oct 2019

Seasonal rain shortages recorded to October 2019 are extreme (1st percentile) in the winters of both 2018 and 2019.
For durations of 9 months up to 96 months, extreme shortages now occur at all of the durations shown . This is much worse than in May. At that time, shortages had not been “extreme” (red), but merely “severe” (grey) at the durations of 36-, 42-, 48-, 60-, and 96-months.
By May, “severe” shortages had already appeared at the very long durations of 120-, 150-, and 240-months. Now, there are also severe shortages at 108-months and 180-months.
The shortage at 240-months has now also become extreme (red). Very low rainfalls since May have dragged down the 240-month total to make it the sixth driest on record at Manilla. This links the short extreme drought of 2002  to become a part of the current drought. That would have seemed unlikely during later deluges, as in summer 2011-12.

Complete Manilla drought record to Oct 2019

Compete record updated to October

When the graph of the complete record of months of rainfall shortage at Manilla is updated to October 2019 it is clear that the current drought is one of the greatest droughts in history.
Now that an extreme rainfall shortage has appeared at 240-month (20-year) duration, along with a complete suite of extreme shortages from 2-months to 96-months, only the droughts of 1912 and 1946 are comparable. The Federation drought of 1902 may also have had a 20-year extreme shortage, but data for that duration are incomplete.


About drought duration graphs

These graphs show the onset, persistence, and breaking of episodes of extreme and severe rainfall shortage (droughts) at Manilla. The Continue reading

October: 20-year extreme shortage

Rainfall status Manilla Sep-Oct 2019

Most rainfall shortages are now extreme

The Rainfall Status graph for October 2019 at Manilla shows extreme rainfall shortages (below the 1st percentile) at most durations.
Of the 25 durations shown, 16 are now extreme shortages, and 7 are severe shortages (below the 5th percentile).
That leaves only 2 that are not far below normal: the October 1-month rainfall (21.4 mm) at the 16th percentile, and the 360-month (30-year) total (18696 mm) at the 14th percentile. Even this 30-year total is lower than any seen here since 1952.

A long-duration extreme shortage

An extreme shortage has now appeared at the very long duration of 240 months (20 years). In the 20 years since November 1999 the total rainfall was only 11893 mm – the 6th driest in history. Drier 20-year periods occurred only in the 1940’s and 1950’s.
Given that the median 20-year rainfall at Manilla is 13010 mm, this is a shortfall of 1117 mm, which is nearly two years of rainfall lost.

Only three new records this month

In this drought, rainfall totals in the 1- to 7-year range have broken records for low rainfall repeatedly. This month only three records have been broken: 24-months (615 mm), 30 months (850 mm), and 36-months (1111 mm).

How to read the graph

This graph shows all the present rainfall shortages at Manilla, short term and long term, as percentile values. The latest values, as at the end of October 2019, are shown by a thick black line with large circles. Those from one month earlier are shown by a thinner line with small diamonds. [The method is described in “Further Explanation” below.]


Further Explanation

The following notes explain aspects of this work under these listed headings:

Data analysis

Cumulative rainfall totals
Percentile values
Severity of rainfall shortages

Limitations of this analysis

Monthly rainfalls form a single population
Observations are not retrospective
The rain gauge failed

Data analysis

Continue reading